The Research On Prediction System Of POEMS Syndrome Based On WeChat Public Platform

POEMS(Polyneuropathy Organomegaly Endocrinopathy Monoclonal protein Skin changes)syndrome is a group of clinical syndromes with multiple sensory motor peripheral neuropathy as the main mianfestation,and frequently accompanied by multiple system damage and clinical syndrome associated with plasma cell abnormalities.It has the characteristics of multiple peripheral neuropathy,organ enlargement,endocrine disorder,M albuminemia and skin disease.POEMS syndrome is a rare disease in the world with an incidence of less than one in a million,but the rate of misdiagnosis and disability is extremely high.WeChat is a free application independently developed by Tencent,and currently has more than one billion active users per month.And has started to provide smart medical services in 2014,and WeChat has been accepted and used by more and more users.Based on this,I designed a POEMS syndrome prediction system based on the WeChat public platform.This paper proposes to establish an information collection platform for POEMS syndrome on the WeChat public platform,which allows users to fill in daily disease records in official accounts at any time,and use XGBoost algorithms to predict POEMS syndrome,so that users enable to record more disease-related data in the official account and provide more clinical data for the study ofrare diseases in the world.In total,the WeChat official account has designed three first-level menus with different functions.Three two-level menus under "POEMS syndrome" were designed by using HTML,CSS,JavaScript and other web page techniques to show the introduction,diagnostic criteria and effective treatment of POEMS syndrome.Three secondary menus under "disease records" for submitting users' daily disease records,uploading medical checklist and viewing historical disease records;The three secondary menus under "personal Information" are used to log in or register your own account,view user information and submit suggestions for WeChat official accounts.The data of 50 different characteristics of 166 patients with POEMS syndrome were collected and divided into training data and test data.After data pre-processing of the training data,the XGBoost algorithm is used to set up a POEMS syndrome prediction model,and the accuracy of the prediction model,the AUC score and the distribution map of the characteristic importance related to the prevalence of POEMS syndrome could be obtained by calculating the accuracy of the prediction model and the score of AUC.The POEMS syndrome prediction system based on WeChat public platform designed and developed in this paper has obvious advantages over the current hospital medical data system.The data of the clinical diagnosis patient in the hospital is closed,and the data among the hospitals are independent of each other.the medical data of this WeChat public platform is open source,the data is more centralized,and it is convenient for medical research.Compared with the hospital,the WeChat public platform is simple and convenient for users,and can also observe the whole process of disease in the longitudinal direction.Therefore,the design of this paper can provide a large number of research data sources for the follow-up POEMS syndrome,and also provide a solution for the study of other rare diseases in the world.